8/4/2019 Mayer, Grimm - Counter Cyclical Fiscal Policy, Simple Rules and Price Dispersion
1/26
Countercyclical Fiscal Policy, Simple Rules
and Price Dispersion
Eric MayerUniversity of Wurzburg
Department of Economics
Sanderring 2
D-97070 Wurzburg, Germany
Oliver GrimmCenter of Economic Reserach
at ETH Zurich
ZUE D12
CH-8092 Zurich, Switzerland
This version: February 15, 2008
Abstract
In this paper we explore the benefits of a supply side oriented fiscal tax policy that
remarkably reduces the impact of nominal frictions on an efficient allocation. We
set up a calibrated version of a New Keynesian DSGE model augmented by a simple
tax rule that engineers an efficient path for the evolution of marginal cost at the firmlevel and prevents any inefficient built up of price dispersion across firms. Taxes
are levied on value added. We highlight that such a policy can be implemented
by simple rules in contrast to optimal commitment solutions and are also effective
under a balanced budget regime. This disencumbers monetary policy in the light
of cost push shocks.
JEL-classification: E32, E61, F62
Keywords: Countercyclical fiscal policy, welfare costs, nominal rigidities
We thank the audience of the German Economic Association Conference in Munich 2007, and of theErich Schneider Seminar 2008 in Kiel for helpful comments.
1
8/4/2019 Mayer, Grimm - Counter Cyclical Fiscal Policy, Simple Rules and Price Dispersion
2/26
1 Introduction
Can fiscal policy eliminate the welfare costs of nominal frictions by means of simple tax
rules? We address this question by setting up a calibrated version of a New Keynesian
DSGE model augmented by a simple tax rule that builds on a rich strand of literature
which has stressed the role of monetary policy to enhance welfare in an environment of
nominal rigidities (Woodford, 2003). However this strand of literature has paid so far
little attention to the role of fiscal policy. In particular, only few studies analyze the
role of distortionary taxation and debt financed expenditures and their implications for
nominal frictions. In this respect, the paper aspires to extend the existing literature in
proposing simple fiscal policy rules that eliminate price dispersion. The government sector
is assumed to maximize the lifetime utility of a representative agent. As instrument it
relies on a value-added tax, debt or government expenditures. Our framework shares most
of the features of recent dynamic optimization sticky price models as e.g. in Woodford
(2003), and Gali, Lopez-Salido and Valles (2007).
In the basic New Keynesian framework there is no reason for output to be different
across firms, except as a result of price distortions that accrue from staggered price setting
(Woodford, 2003). Through this channel infrequent price adjustments create undesired
variations in the relative prices of goods across firms. Under standard assumptions on
preferences this creates welfare losses on the consumption side as households buy more
of the relatively cheaper goods and less of the more expensive ones. On the production
side the increased cost of producing some goods more is greater than the cost saving of
producing less of other goods. Through these channels price stability turns out to be
a dominant target of monetary policy (Woodford 2003, p. 396). A sufficiently strong
feedback from movements in the inflation rate is argued to be the best response to limit
the adverse effects of cost-push shocks on lifetime utility of a representative consumer.
Notwithstanding the previous argument the welfare costs of nominal rigidities are esti-
mated to be up to three percent in consumption equivalents (Canzoneri, Cumby and Diba,2007; and Gertler and Lopez-Salido, 2007).
This highlights that monetary policy does not have a direct leverage on the supply side
and thus the price setting behavior of firms. Monetary authorities can only control price
dispersion through the aggregate demand channel and thus the reallocation of intertem-
poral consumption plans. Therefore, in the event of a supply shock firms are tempted
to increase prices. As best response monetary authorities raise the real interest rate to
encourage consumers to reallocate consumption to the future which depresses contempo-
2
8/4/2019 Mayer, Grimm - Counter Cyclical Fiscal Policy, Simple Rules and Price Dispersion
3/26
raneous demand and thus demand driven production plans. As production plans have
to be consistent with the labor supply schedules of workers an equilibrium only occurs if
wages and thus marginal cost decline. This stands in contrast to demand shocks which
can be wiped out at zero cost (Clarida, Gali, Gertler, 1999). In essence it is the lack ofan additional instrument on the supply side of the economy on part of monetary author-
ities which makes markup shocks costly in terms of welfare. To that extend we follow
the Tinbergen logic and propose that fiscal policy should use its value-added tax as an
additional instrument in a state contingent way such that the evolution of marginal cost
is stabilized around its deterministic steady state (Tinbergen, 1959). Those firms that are
called upon to reset prices will then built on the promise of fiscal authorities to smooth
away cost push shocks and set prices in the neighborhood of those price setters that have
to leave prices unchanged.A key finding of our paper is that fiscal authorities can target a path for value-added
taxes that evolves countercyclical to markup shocks and thus prevents large movements
in marginal cost. This seems in particular important as Schmidt-Grohe and Uribe (2006)
report evidence from a medium-scale model which comprises a number of real and nominal
frictions that price stickiness emerges as the most important distortion.
When fiscal policy is allowed to cushion changes in tax rates by debt rather than gov-
ernment expenditures we report evidence that debt prevails a near-random walk behavior
in the presence of cost-push shocks. Schmitt-Grohe and Uribe (1997) pointed out thatdistortionary taxes might produce multiple equilibrium if they are conditioned on the
state of the cycle. Notwithstanding these results it has prevailed in DSGE models that
multiple equilibrium can be ruled out if fiscal authorities react sufficiently strong with
the tax rate to changes in the level of existing debt (Canzoneri, Cumbi and Diba, 2003;
Linnemann and Schabert, 2003). In particular the steady state levels of tax rates and
government expenditure is determined by long-run solvency considerations such that in
steady state the budget is balanced.
Although the general idea of simple fiscal rules is not new previous authors mainly
focused on the idea of classical demand management, where government expenditures are
conditioned on the output gap such as J.B. Taylor (2000). Only few studies consider
stochastic taxation, and its role for nominal frictions. A notable exception is Leith and
Wren-Lewis (2007), who explore the role of countercyclical fiscal policy in a full-fledged
DSGE model and analyze commitment solutions. They report evidence that price dis-
persion can be completely wiped out by commitment solutions when fiscal authorities
employ three instruments, namely debt, government expenditures and tax rates on labor
3
8/4/2019 Mayer, Grimm - Counter Cyclical Fiscal Policy, Simple Rules and Price Dispersion
4/26
and value added. We differ from there work in several aspects: (i) instead of modeling
commitment solutions we show that simple rules are sufficient to substantially improve
welfare. (ii) We report evidence that such rules are also effective under a balanced-budget
regime by means of MSV-solutions. (iii) We obtain results from a sensitivity analysiswith respect to deep parameters. (iv) We simulate the behavior of the economy with the
occurrence of markup shocks.
Our findings suggest that countercyclical supply-side taxation rules can remarkably
reduce the impact of cost-push shocks on welfare by reducing the welfare loss substantially.
The paper is structured as follows: In Section 2, the basic model is introduced. Section 3
presents analytical results on simple rules and price dispersion. In Section 4 we exhibit the
numerical findings. In Section 5 we conduct robustness analysis. Section 6 summarizes
the main findings and concludes.
2 The Model
In this section we present a New Keynesian DSGE model that consists of firms, house-
holds, the central bank and fiscal authorities. As standard, firms are partitioned into the
final good sector and a continuum of intermediate good producers. Intermediate good
producers have some monopoly power over prices that are set in a staggered way fol-
lowing Calvo (1983). Households obtain utility from consumption, leisure and invest in
state contingent securities. Monetary authorities are guided by a simple Taylor rule. The
government sector is financed by distortionary taxes levied on value added or debt. Fiscal
policy is implemented by simple tax and spending rules.
The model builts on the framework of Gali, Lopez-Salido and Valles (2007), Leith
and Wren-Lewis (2007), and Linnemann and Schabert (2003) by sharing the same kind
of features such as debt financed expenditures, state contingent tax rules and staggered
price setting as in Calvo (1983). In particular we highlight the role of an active fiscal
policy compared to a neutral stance to fight the welfare costs of price dispersion
2.1 Final Good Producers
The final good is bundled by a representative firm which operates under perfect compe-
tition. The technology available to the firm is:
Yt =
1
0
Xt(i)1 di
1
, (1)
4
8/4/2019 Mayer, Grimm - Counter Cyclical Fiscal Policy, Simple Rules and Price Dispersion
5/26
where Yt is the final good, Xt(i) are the quantities of the intermediate goods, indexed
by i (0, 1) and > 1 is the elasticity of substitution. Profit maximization implies the
following demand schedules for all i (0, 1):
Xt(i) =Pt(i)
Pt
Yt. (2)
The zero-profit theorem implies Pt =
1
0Pt(i)
1 1
1, where Pt(i) is the price of the
intermediate good i (0, 1).
2.2 Intermediate Good Producers
Firms indexed by i (0, 1) operate in an environment of monopolistic competition. The
typical production technology is given by:
Yt(i) = Nt(i) , (3)
where Nt(i) denotes labor services. Nominal profits by firm i are given by:
t(i) =
1 V ATt
Pt(i)Yt(i) WtNt(i), (4)
with Yt(i) = Xt(i) and V ATt denotes a value-added tax with
V ATt (0, 1). As cost
minimization implies that marginal costs are equal to wages with t = wt the profit
function can be rewritten as follows:
t(i) =
1 V ATt
Pt(i) Ptt
Yt(i). (5)
The representative firm is assumed to set prices as in Calvo (1983), which implies that
the price level is determined in each period as a weighted average of a fraction of firms
(1 p) which resets prices and a fraction of firms p that leaves prices unchanged:
Pt = (1 p)(Pt)1 + pP
1t1
11
, (6)
where Pt is the optimal reset price in period t. Profit maximization comprises the following
first-order condition:
Et
k=0
(p)kt+kYt+k(i)
Pt(i)(1
V ATt ) t+kPt+kt+k(i)
, (7)
where t denotes the stochastic discount factor of shareholders to whom profits are re-
deemed and t denotes the gross frictional markup prevailing in an environment of flexible
prices and zero inflation. The deterministic counterpart of t is ( 1)1 .
5
8/4/2019 Mayer, Grimm - Counter Cyclical Fiscal Policy, Simple Rules and Price Dispersion
6/26
2.3 Households
We assume a continuum of households indexed by j (0, 1). A typical household seeks
to maximize lifetime utility:
E0
t=0
tUt(j) , (8)
where denotes a discount factor with (0, 1), and period utility is given by:
Ut(j) = (1 )
1
1 Ct(j)
1
+ Gt
1
1 + Nt(j)
1+ . (9)
is a coefficient of risk aversion, is the inverse of the Frisch elasticity of labor supply. Ct
are the real consumption expenditures of household j. The sequence of budget constraints
reads:Ct(j) +
Bt+1(j)
RtPt
WtNt(j)
Pt+
t(j)
Pt+
Bt(j)
Pt. (10)
Each household decides on consumption expenditures Ct(j) and bond holdings Bt+1(j)
and receives labor income WtNt(j) , dividends from profits t(j)/Pt and the gross return
on bonds purchased Bt(j).
Maximizing the objective function subject to the intertemporal budget constraint with
respect to consumption and bond holdings delivers the following first-order conditions:
(1 )C
t = t, (11)
Nt = wt (12)
1
Ptt = Et
t+1
1
Pt+1Rt
, (13)
where t denotes the Lagrangian from relaxing the budget constraint. Conbining the first
order conditions yields the consumption Euler equation and the labor supply schedule:
Ct = RtEt
Ct+1
PtPt+1
(14)
N
t
1 Ct = WtPt
. (15)
2.4 Fiscal Authorities
The government issues bonds and collects value-added taxes. It uses its receipts either to
finance government expenditures or interest on outstanding debt. The real government
budget constraint reads:
Bt+1 + V ATt Yt = RtBt + Gt . (16)
6
8/4/2019 Mayer, Grimm - Counter Cyclical Fiscal Policy, Simple Rules and Price Dispersion
7/26
Letting bt =1
Y[(Bt/Pt1) (B/P)] and B = 0, the budget constraint can be rewritten as:
bt+1 + V ARt
YtY
= RtbtPt1
Pt+
GtY
. (17)
Government purchases are assumed to move countercyclical to output:
Gt = YoY
t , (18)
where oY < 0 denotes the expenditure elasticity with respect to income Yt. The tax rule
reads:
V ATt = 1t b
2t+1 , (19)
which is conditioned on the stochastic markup shock t and outstanding debt bt+1. In
principle a sufficient strong response to the change of the level of outstanding debt 2 > 0
assures uniqueness and determinacy. A parameter 1 < 0 denotes a countercyclical fiscal
tax policy.
2.5 Market Clearance
In clearing of factor markets and good markets the following conditions are satisfied:
Yt = Ct + Gt,
Yt(j) = Xt(j),Nt =
1
0
Nt(j)dj.
2.6 Linearized Equilibrium Conditions
In this section we summarize the model by taking a log-linear approximation of the key
equations around a symmetric equilibrium steady state with zero inflation and zero debt.
A variable Xt denotes in the following the log-linear deviation from the steady state value:
Xt
= log(Xt) log(X), where X represents the steady state.
Households The consumption Euler equation reads:
Ct = EtCt+1 1(Rt Ett+1) , (20)
where t is defined as t Pt Pt1. Under perfectly competitive labor markets the
labor supply schedule is equal to:
wt = Nt + Ct . (21)
7
8/4/2019 Mayer, Grimm - Counter Cyclical Fiscal Policy, Simple Rules and Price Dispersion
8/26
Firms Log-linearization of (6) and (7) around a zero inflation steady state yields the
dynamics of inflation as a function of the wage wt, a stochastic markup t and tax rates
V ATt :
t = t+1 + [ wt + V ATt + t], with
(1 p)(1 p)
p, (22)
where /(1 V AT).
Fiscal authorities Log-linearizing the budget constraint around a zero steady state
debt yields the following approximation up to first order:
bt+1 + G(V ATt + Yt) =
1bt + GGt , (23)
where for the case of a balanced budget (23) simplifies to Gt = V ATt + Yt. The parameter
G denotes the steady state government share which is equal to V AT implied by a balanced
budget in steady state. The fiscal spending rule is the log-linearized version of (18):1
Gt = oY Yt . (24)
The simple tax rule is the log-linearized complement to (19):
V ATt = 1t + 2bt+1, (25)
In the following we will refer to a passive fiscal policy if 1 = 0 which comprises that
movements in tax revenues V ATt + Yt are passively absorbed by movements in debt bt+1
and changes in the tax rate V ATt = 2bt+1.
Monetary Policy Monetary policy is assumed to follow the Taylor rule:
Rt = (1 )Rt1 + [ + t + xxt] , (26)
where and x capture the reaction coefficients with respect to the inflation rate and the
output gap xt as defined below. The rule satisfies the Taylor-principle as long as > 1,
1Note that the welfare criterion is derived for oY = 0.
8
8/4/2019 Mayer, Grimm - Counter Cyclical Fiscal Policy, Simple Rules and Price Dispersion
9/26
which is a necessary requirement for uniqueness and stability (Woodford, 2003).
Market Clearing Market clearing requires that the following relation holds:
Yt = CCt + GGt , (27)
where C denotes the consumption share, which is equal to (1 V AT). Using (24) and
(27) we can rewrite the consumption Euler-equation as follows:
Yt = EtYt+1 CRt Ett+1(1 GoY)
. (28)
Flex-price equilibriumThe flex-price equilibrium is obtained by equating wt =
Nt + Ct and t = wt which combines the real marginal product of labor to the marginal
rate of substitution between consumption and leisure:
ft = Yf
t , with [ + 1
C (1 GoY)] , (29)
where we additionally used the fiscal spending rule (24) and market clearance condition
(27). The superscript f denotes flexible prices. From the optimal price-setting behavior
of firms operating in the intermediate good sector under flexible-prices we know that:
ft = 1t (1 V ATt ) , (30)
where we assumed that fiscal policy sets 1 = 0 if prices are flexible as no price dispersion
prevails in the flex-price equilibrium such that V AT,ft = 2bft+1. Accordingly the log-
deviation of real marginal cost from its deterministic counterpart ( 1)/ can then be
written in log-linearized terms as: ft = (ttV AT,f). If we define xt = Yt Y
ft as the
welfare gap, i.e. the gap between the actual output gap and the output gap under flexible
prices the log-deviation of marginal cost can be written as:
t = (xt + Yft ), with Yft = 1 (t + V AT,ft ) . (31)
We can rewrite the Phillips curve in terms of xt as:
t = Ett+1 + [xt + (V ATt
V AT,ft )] , (32)
From the Euler-equation we know that the natural rate of interest under flexible prices is
equal to:
rnt = (EtYf
t+1 Gft+1) , (33)
9
8/4/2019 Mayer, Grimm - Counter Cyclical Fiscal Policy, Simple Rules and Price Dispersion
10/26
where log . Inserting Yft+1 and Gft+1 the natural rate can be expressed in terms
of the exogenous shock t+1 and the tax rule V AT,ft under flexible prices:
rnt = (1 oY )1
[Ett+1 + EtV AT,ft ] . (34)
Using the definitions of the welfare gap xt it holds that:
xt = Etxt+1 C((1 Goy))1[Rt Ett+1 r
nt ] , (35)
and
bt+1 GV AT,ft =
1bt + G(oY 1)xt + Gt GV ATt , (36)
with 1
(1 oY).
Discussion Notwithstanding that most of the features in the model are standard in
particular the value-added tax augmented Phillips curve is worth stressing. First, notice
that the inflation rate is simply a weighted average of the expected path of wage costs,
the markup shock and the evolution of the value-added taxes. As we will show below this
enables the government to design a path for value-added taxes which almost completely
offsets any movement in cost pressure such that price dispersion across firms can be
reduced. Secondly, as we formulate state contingent tax and spending rules government
debt necessarily works as a buffer to accommodate movements in the spending rule and
movements of the tax rate. For the case of a balanced budget regime movements in the
tax rate call for adjustments in fiscal spending.
3 Simple Rules and Price Dispersion
In this Section we analytically examine the role of simple tax rules on the equilibrium
allocation of inflation, output, consumption, interest rates and government expenditures.
To keep the calculations analytically tractable, we assume that the budget is balanced such
that (25) reduces to V ATt = 1t and government expenditures are adjusted passively
such that the budget equation (23) holds. Additionally, we reduce the system by inserting
the natural rate of interest rnt and the tax rule (25) into the Phillips curve (32) and the
Euler-equation (35). Then the model can be written as the following set of expectational
10
8/4/2019 Mayer, Grimm - Counter Cyclical Fiscal Policy, Simple Rules and Price Dispersion
11/26
difference equations:
xt = Etxt+1 1(Rt Ett+1) + (G
1
C 1 + ( + )1)t (37)
t = Ett+1 + [( + )xt + (1 )G1
C 1t] (38)
Rt = t (39)
where the coefficient 1 serves as a parameter which can be freely chosen by fiscal author-
ities. The following propositions summarize the main results.2
Proposition 3.1 Suppose that a social planer is only concerned about price disper-
sion. Then choosing a coefficient 1 = C1
G (1 + )1 completely eliminates any price
dispersion across firms at any date t.
Proof Since the simplified model with bt+1 = bt = 0 exhibits no endogenous state
variables the fundamental solution takes the form: t = t. Applying the methods of
undetermined coefficients leads to the following solution: = [1 + ( + )1]
1[1 +
G1
C 1(1 + )]. Inflation is completely stabilized if = 0 which holds for 1 =
C1
G (1 + )1.
Thus according to Proposition 3.1 fiscal authorities can completely stabilize the inflationrate by choosing 1 appropriately. For the applied calibration, 1 would take a numerical
value of 1 = 2.00 (G = 0.2; = 1). Interestingly the coefficient 1 only depends
on two deep parameters, namely V AT and . In line with intuition an increasing steady
state government share G increases the leverage of fiscal authorities on real marginal
costs and on prices such that the same effects on equilibrium allocations can be achieved
by smaller movements of the instrument V ATt . Additionally, the responsiveness of the
coefficient 1 decreases in the Frisch elasticity of labor supply. Thus, if labor supply is
more responsive to wage changes equilibrium wages will be less responsive over the cycle
such that smaller movements in the tax instrument are sufficient to yield the same effects
on the evolution of marginal cost.
Proposition 3.2 Suppose that we compare two economies which are identical except
that in one economy fiscal policy implements the simple rule V ATt = 1t whereas in the
2For the MSV-solutions, see appendix B.
11
8/4/2019 Mayer, Grimm - Counter Cyclical Fiscal Policy, Simple Rules and Price Dispersion
12/26
other economy fiscal policy remains passive with V AT = V ATt and G = Gt t. Then for
any policy choice with 1 < 0 the evolution of the inflation rate t, the welfare gap xt
and nominal interest rates Rt evolve smoother than in an economy where 1 = 0.
Proof Since in both economies the simplified model exhibits no endogenous state vari-
able the fundamental solution takes in both cases the following form, with Xt = X t,
with Xt = [t xt Rt] and X = [ x R]. Thus a necessary and sufficient condition for
a smoother evolution of the economy is |AX |i,1 < |PX |i,1 for i = 1, 2, 3, where the super-
scripts A denote active and P passive. As shown in appendix B a necessary and sufficient
condition for this inequality to hold is that 1 < 0.
Thus according to Proposition 3.2 it holds that any policy choice with 1 < 0 accommo-
dates a smoother evolution of the economy.
Without any statement on welfare, we can already conjecture that an active fiscal
stance is welfare improving if government expenditure is pure waste as the welfare function
for this case would only built on the inflation rate t and the welfare gap xt. Note that
we know by the Taylor Rule that the nominal interest rate will be smoothed as it is just
a linear transformation of the inflation rate itself. This in turn, implies that a smoother
evolution of the real interest rate fosters a stable consumption path (Ct/Ct+1) as can be
seen from the Euler-equation:
1
= Et
Ct
Ct+1
Rt
t+1
, (40)
which states that the product of the logdeviation of the real interest rate and the ratio
of the transformed logdeviations of consumption will always be equal to the inverse of
the discount factor. For the case of a balanced budget changes in the tax rate have
to be cushioned by fiscal spending. Therefore fiscal spending Gt is by definition more
volatile than under a passive fiscal stance. Notwithstanding the argument the output
gap Yt, defined as the weighted sum CCt + GGt, evolves less volatile. This reflects
that the additional volatility in government expenditure is overcompensated by the stable
evolution of consumption, which implies that such a policy is welfare improving as long
as consumers attach a higher weight to consumption than to government expenditures in
their welfare metric.
12
8/4/2019 Mayer, Grimm - Counter Cyclical Fiscal Policy, Simple Rules and Price Dispersion
13/26
4 Welfare
Next we characterize the model if we allow for debt financed expenditures by means of
numerical analysis. As shown in the appendix C a second-order approximation around
a steady state to the average utility of a household can be written as follows (see Erceg,
Henderson, and Levine, 2000, Gali and Monacelli, 2007, and Woodford 2003).
Wt =
t=0
tLt , (41)
where
Lt =
2t + (1 + )Y
2
t + (Gt Yt)2. (42)
In the following we discuss the implementation of the proposed tax rule (25). Since we
do not have a distinctive imagination for appropriate numerical parameters except that
1 < 0 and that 2 > 0, we opt to choose the parameters such that the welfare function
(41) is minimized. This does not mean that we want to propagate simple optimized rules
but it simply means that due to the lack of parameters to calibrate this seems to be the
most reasonable way to proceed as a first guess.
Figure 1 portrays the dynamic responses of selected variables to a markup shock.
For the baseline case fiscal policy remains passive with 1 = 0 whereas for the active
stance with 1 < 0 fiscal policy aspires to improve welfare by controlling the evolution of
marginal cost. The following remark summarizes the main findings:
Remark: The implementation of rule (25) largely disconnects the evolution of the infla-
tion rate from exogenous markup shocks. If free to choose fiscal authorities prefer long
debt cycle to cushion the exogenous shock.
The impulse responses portray that a sharp cut in taxes V ATt levied on the value-added
prevents any built up in cost pressure. The tax cut occurs in particular in the first
and second quarter, when the geometrically decaying markup shock hits strongest. As a
fraction of firms P is called upon to reset prices they foresee that any price pressure is
undone by fiscal authorities by the targeted tax path that keeps the sum of wage path,
markup shock and tax path flat. Due to the moderate evolution of the inflation rate
monetary authorities are prevented from sharply raising nominal interest rates. This in
turn detains Ricardian households to reallocate planned consumption expenditures by
13
8/4/2019 Mayer, Grimm - Counter Cyclical Fiscal Policy, Simple Rules and Price Dispersion
14/26
large into the future. As consumption accounts for 80% of output we observe a moderate
drop in production. If fiscal authorities are free to choose, they absorb the tax cut by a
near-random walk behavior in debt. Note as markup shocks are symmetrically distributed
a near-random walk behavior in debt implies that the persistent swings cancel out eachother. Contemporaneous government-expenditure changes on the contrary are welfare
reducing as they increase the expected variability in government expenditures. For the
Figure 1: Stabilization by Fiscal Tax Rules
0 4 8
0
0.1
0.2
t
0 4 81
0.5
0
Yt
0 4 81
0.5
0
Ct
0 4 80.2
0
0.2
0.4
Rt
0 4 81.5
1
0.5
0
wt
0 4 8
0
0.5
1
rf
t
0 4 8
4
2
0
t
0 4 80
0.2
0.4
Cost pressure
0 4 8
1
0
1
2
3
Bt+1
1=0
10
Notes: Responses of selected variables to a markup shock. Solid lines indicate a state indepen-dent passive fiscal policy with 1 = 0 . The dashed-dotted line shows the impulses of the model
when fiscal policy is active with 1 < 0 and 2 > 0 . For the applied baseline calibration seeappendix A.
baseline scenario the simulations indicate that the implementation of the simple rule
reduces the value of the loss function by 62 percent. The numerical values are given by
1 = 0.06 and 2 = 1.64. Note that the numerical value is close to the one found in the
analytical section of the paper.
14
8/4/2019 Mayer, Grimm - Counter Cyclical Fiscal Policy, Simple Rules and Price Dispersion
15/26
5 Relevance of the Tax-Rule
Markup shocks are costly in terms of welfare as monetary authorities lack an instrument
on the supply side of the economy to cushion the adverse effects of cost pressure. Following
the Tinbergen (1959) logic we have shown that a state contingent tax, conditioned on the
state of the markup shock and the level of debt, can improve welfare remarkably.
In the following we discuss the implications of these issues by computing welfare gains
using different parameter constellations. This exercise has two main purposes. On the
one hand we want to analyze whether the proposed rule is robust to perturbations of the
baseline parameterization. On the other hand we present further insights why the rule
works from a micro-founded perspective.
Figure 2: Recalibrating the Baseline Model Loss Ratio
1 40.35
0.4
0.45
0.5
0.55
0.6
0.65
0.7
0.75
LossRatio
1.1 2
0.4
0.5
0.6
0.7
0.8
0.9
1
LossRatio
0 10.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
x
LossRatio
0.50 0.85
0.4
0.5
0.6
0.7
0.8
0.9
1p
LossRatio
0 0.750.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
LossRatio
1
baseline
1
upper
1
lower
0 0.90
0.4
0.5
0.6
0.7
0.8
0.9
1
LossRatio
Notes: Evolution of the expected loss ratio defined as the ratio of the expectedloss if fiscal policy is active with 1 < 0 compared to a passive stance 1 = 0,
E0
t=0 tLt
Active/
t=0 tLt
P assive. Appendix A summarizes the ranges of deep pa-
rameters typically found in the literature. The dashed-dotted and the dotted line indicate theevolution of the loss ratio if the reaction coefficient 1 is either reduced or augmented by fiftypercent from the optimal value.
Precisely speaking we computed the expected value of the loss E0{
t=0 tLt} for the
active and the passive fiscal policy stance and then take the ratio of the two. If the ratio
takes the value one, then the loss would be equal under the two regimes. If the value of the
15
8/4/2019 Mayer, Grimm - Counter Cyclical Fiscal Policy, Simple Rules and Price Dispersion
16/26
ratio is below (above) one, then the loss under an active fiscal policy is smaller (larger)
than the loss under the passive fiscal stance. By means of computing these ratios we
succeed to uncover those parameter constellations which improve or worsen the relative
performance of the proposed policy rule compared to the fallback position of a passivefiscal policy. The solid line indicates how the computed ratio changes when the parameter
displayed at the top of the figure is altered, while the rest remains fixed at the baseline
calibration. For each altered coefficient, e.g. for , the coefficients in the fiscal policy rule
1 and 2 are reoptimized such that (41) is minimized. As a robustness exercise we also
report the ratio if we deviate from 1 according to 1 22
indicated by upper1 and lower2
in Figure 2. Generally the results indicate no large asymmetries when fiscal authorities
tend to choose too high or too low coefficients 1, which indicates that the loss ratio
behaves largely linearily when deviating from the baseline by altering 1. For the case oflarge asymmetries we would have expected the reported lines upper1 and
lower2 to have a
substantial distance from each other.
The inverse of the Frisch elasticity of labor supply was varied one to four. The
robustness analysis indicates that irrespectively of where the parameter is calibrated
the relative advantages of the policy rule remains unaltered by a reduction in relative
losses of round about 62%.
With respect to the Taylor rule coefficient the relative advantage of the fiscal policy
rule increases if monetary policy gets somewhat more aggressive on inflation. This mightreflect to a certain extend that a larger Taylor-rule coefficient implies that the real
interest rate volatility and thus the variations in the consumption aggregate over time
increase. One outstanding effect of the fiscal policy rule (25) is that it disencumbers
monetary policy such that there is no need, even if monetary policy takes an aggressive
stands towards inflation to move the real interest rate instrument a lot.
The robustness analysis indicates that the relative advantage of the proposed policy
rule decreases if monetary policy reacts stronger on the welfare gap. Nevertheless the
relative performance does not deteriorate more than round about twenty percent, even
for a coefficient of x = 1, which is higher than the values typically found in literature.
This might be explained by the fact that the proposed tax rule is successful in reducing
inflation, but not so much in reducing output gap variability. This implies that a monetary
authority that takes the output gap into account reintroduces real interest rate variability.
The performance of the rule worsens if interest rates are set in a highly inertial fashion.
Nevertheless the ratio only deteriorates by 15% when increases from 0 to 0.75. The
effectiveness of the rule increases with the degree of correlation in the markup shock .
16
8/4/2019 Mayer, Grimm - Counter Cyclical Fiscal Policy, Simple Rules and Price Dispersion
17/26
The higher the degree of correlation the larger will be the price dispersion inflicted upon
the economy. Those firms that are called upon to reset prices will anticipate further shocks
in the same direction which triggers a larger adjustment of prices. Therefore, the rule
is in particular welfare enhancing in an environment of correlated shocks as it promisesto firms a stable evolution of prices and thus a limited degree of price dispersion for the
economy.
With respect to the value of the Calvo parameter P there exists a considerable dis-
agreement in the literature. Del Negro et. al. (2005) for instance estimate an average
price duration of three quarters for the euro-area using full information Bayesian tech-
niques; Gali, Gertler, and Lopez-Salido (2001) report a value round about four quarters
using single equation GMM approach. Empirical work on price setting in the euro area
using micro evidence report relatively low price durations with a median round about 3.5quarters (see Alvarez et. al., 2006, for a summary of recent micro evidence). Comparable
studies for the U.S. like Altig et. al. (2005) report much lower average price durations
of just 1.6 quarters, which they claim to be more consistent with recent evidence drawn
from US micro-data. Based on this review of the literature it seems fair to conduct the
robustness analysis in a range between 1.7 to five quarters, which corresponds to P rang-
ing between 0.45 to 0.85. The figure illustrates that the ratio monotonically decreases
from 0.45 to 0.38 when the Calvo parameter is altered from 0.4 to 0.85. Thus, the rule
performs better when the degree of nominal stickiness increases. This reflects that withan increase of nominal frictions a price distribution across firms unfolds if fiscal policy
remains passive. The implementation of the policy rule prevents that a wedge can be
driven between the production schedules and thus enhances welfare as the variability of
inflation decreases. However, for a very high degree of price rigidity (P > 0.80) the ratio
increases, as the fiscal rule loses its effectiveness.
6 Conclusions
In this paper we addressed the question whether fiscal policy can wipe out price dispersion
by implementing a simple tax rule. Our motivation stems from the fact that there is a
large strand of literature which stresses the role of monetary policy to enhance welfare in
an environment of nominal rigidities (Woodford, 2003). However this strand of literature
has paid so far little attention to the question whether fiscal policy can improve welfare
with respect to nominal frictions. In the event of cost push shocks Woodford (2003) shows
that monetary policy faces a trade off between stabilizing the inflation rate and stabilizing
17
8/4/2019 Mayer, Grimm - Counter Cyclical Fiscal Policy, Simple Rules and Price Dispersion
18/26
the output gap. A sufficiently strong feedback from movements in the inflation rate is
argued to be the best response to limit the adverse effects of cost-push shocks on lifetime
utility of a representative consumer to generate a unique and determinate equilibrium.
Notwithstanding, these arguments the costs of nominal rigidities are estimated to be stillup to three percent in consumption equivalents (Canzoneri, Cumby and Diba 2007).
This highlights that monetary policy does not have a direct leverage on the supply
side of the economy. Therefore, we followed the Tinbergen logic and proposed that fiscal
policy should use its value-added tax, as an additional instrument in a state contingent
way such that the evolution of marginal cost is stabilized around its deterministic steady
state. Our findings suggest that a countercyclical taxation approach can remarkably
reduce the impact of cost push shocks on welfare. The reduction in expected losses,
when fiscal authorities switch from a passive towards an active fiscal stance are quantifiedaround about fifty percent and depend on the particular parameter settings. Key to the
functioning of the tax-rule is that the fraction of firms that adjusts prices anticipates
the promise of fiscal authorities to target a value-added tax path that eliminates any cost
pressure at the firm level. Accordingly, those firms that are called upon to reset prices will
set them in the neighborhood of those firms that leave prices unchanged. This prevents
any inefficient built up in prices across firms at any date t.
The Keynesian tradition considers fiscal policy as operating over the aggregate de-
mand effect. We showed that fiscal policy can use its distortionary instruments to unfoldstabilizing effects on the economy upon an aggregate supply channel.
18
8/4/2019 Mayer, Grimm - Counter Cyclical Fiscal Policy, Simple Rules and Price Dispersion
19/26
Appendices
A Calibrated Parameters
In Section 5 of the main text we conduct some sensitivity analysis to demonstrate the
robustness of the proposed policy rule. While conducting this exercise we rely on ranges
of the deep parameters chosen in a way to best represent the uncertainty found in the
literature as reported in Table 1.
Table 1: Values and Ranges for the Calibrated Parameters
Parameter Baseline RangeA. Household
Discount factor 0.99 /
Risk Aversion 1.00 1.00 4.00
Inverse of the Labor Supply Elasticity 2.00 1.00 4.00
B. Firms
Price Elasticity of Demand for an Intermediate Good
Variety
6.00 6.00 25.00
Price Stickiness P 0.75 0.40 0.85C. Monetary Policy
Taylor Rule: Smoothing 0.50 0.00 0.75
Taylor Rule: Inflation 1.50 1.10 2.00
Taylor Rule: Welfare Gap x 0.25 0.00 1.00
D. Fiscal Authorities
Fiscal Rule: Mark-up shock 1 -1.64 /
Fiscal Rule: Debt 2 0.06 /
Gross Steady State Tax Share V AT 0.20 0.10 0.50E. Exogenous Shock
Mark-up Shock: Persistence 0.75 0.00 0.90
Remarks: The table displays the calibrated values. The respective upper and lower boundsare taken from related studies in literature. The reviewed literature is Smets and Wouters,2003; Leith and Maley, 2005, Rabanal, 2003, Coenen, McAdam and Straub, 2006, Del Negro,Schorfheide, Smets and Wouters, 2004, Welz, 2005.
19
8/4/2019 Mayer, Grimm - Counter Cyclical Fiscal Policy, Simple Rules and Price Dispersion
20/26
B Derivation of the MSV Solution
Balanced budget and active stance Substituting out the tax-rate V ATt and the
natural rate rnt of interest the reduced form system can be written as:
xt = Etxt+1 1(t Ett+1) +
GC
1 + ( + )1
t , (B.1)
t = Ett+1 +
( + )xt +
(1 )
GC
1
t
, (B.2)
t = t (B.3)
The rest of the system is recursive and can be solved afterwards. Let us posit a funda-
mental (minimum state variable) solution of the following generic form (McCallum, 1983):
t = t and xt = xt, where the coefficients and x remain to be determined. WithEtxt+1 = Etxt = 0 and Ett = Ett+1, this leads to the following conditions for the
undetermined coefficients:
= ( + )x + (1 )GC
1 , (B.4)
x = 1 + ( + )
1 +GC
1 . (B.5)
Inserting (B.5) into (B.4) yields
= [1 + ( + )1]
1 [1 + G1C 1(1 + )], (B.6)
and
x =1
+
1
1 + 1( + )+
GC
1 + (1
1)
1 + 1( + )1 . (B.7)
Balanced budget and passive policy Let us define the neutral benchmark system
as Gt = V ATt = 0. Then the model can be stated as:
xt = Etxt+1 1
(t Ett+1) + ( + )1
t , (B.8)t = Ett+1 + ( + )xt , (B.9)
where the MSV solution reads:
x = [1 + 1( + )]
1( + )1 , (B.10)
and
= [1 + 1( + )]
1 . (B.11)
20
8/4/2019 Mayer, Grimm - Counter Cyclical Fiscal Policy, Simple Rules and Price Dispersion
21/26
Comparison of active versus passive fiscal policy In the following we compare
the MSV solutions for an economy where fiscal policy implements policy rule (24) versus
an economy where fiscal policy remains passive with Gt = V ATt = 0. The superscript P
denotes passive whereas the superscript A denotes active.
Inflation:
P > A
[1 + 1( + )]1 > [1 + 1( + )]
1[1 + G1
C 1(1 + )]
1 > 1 + G1
C 1(1 + )
0 > G1
C (1 + )1 1 < 0, , G, C > 0
Welfare gap:
Px > Ax
[1 + 1( + )]1( + )1
> [1 + 1( + )]1( + )1 [1 + G
1
C (1 + (1 1))1]
1 > 1 + G1
C (1 + (1 1))1
0 > G1
C (1 + (1 1
))1 1 < 0, G, C, , , > 0
C Utility-Based Welfare Function
The utility function is given by:
U(C , N, G) = (1 )log C+ log G N1+
1 +
. (C.1)
Note that the weight in the utility function is equal to the steady state share of govern-
ment spending = G/Y. Taking a second-order approximation around the consumption
part of the utility function yields:
log(Ct) = log(Yt Gt) =1
1 (yt g)
1
2
(1 )2(yt gt)
2 + tip + o(||a3||) . (C.2)
Where it holds that: xt = xt + (xt x). We denote the gap yt = yt yt and the fiscal
gap gt = gt gt. Note that yt comprises the sum of the deviation of output from the
21
8/4/2019 Mayer, Grimm - Counter Cyclical Fiscal Policy, Simple Rules and Price Dispersion
22/26
distorted (short term) steady state and the deviation of the distorted steady-state output
from the efficient long-term steady state. Taking a second-order approximation around
the disutility of labor term yields:
N1+
1 + = nt +
12
n2t + tip + o(||a3||). (C.3)
We find the relationship Nt = YtQt, which is derived in the following:
Nt =
1
0
Nt(i)di =
1
0
Yt(i)di = Yt
1
0
Yt(i)
Yt(C.4)
Nt = Yt
1
0
Pt(i)
Pt
di
Qt= YtQt . (C.5)
After log linearization, we obtain:
nt = yt + qt . (C.6)
Where qt = (/2)2t and qt is defined as:
qt log
1
0
Pt(i)
Pt
di . (C.7)
The intertemporal welfare function is given by the discounted sum of the approximated
utility functions:
Wt =
t=0
tUt(Ct, Gt, Nt) =
t=0
t
(1 + )y2t + (gt yt)2 + 2t
. (C.8)
Now, we aim at expressing 2t in terms of2t while following the proof given by Woodford
2003:
t=0 t2t =
t=0 t tip +
t
s=0 tsP
P
1 P2s + o(||a||
3)=
1
t=0
t2t + tip + o(||a||3) . (C.9)
Using this result (C.8) can be rewritten as follows:
Wt =
t=0
t
2t + (1 + )y
2
t + (gt yt)2
. (C.10)
22
8/4/2019 Mayer, Grimm - Counter Cyclical Fiscal Policy, Simple Rules and Price Dispersion
23/26
D Matrix Notation of the Model
The linearized equilibrium dynamics can be represented as follows (Soderlind, 1999):
A0 X1,t+1
EtX1,t+2
= A1
X1,tEtX1,t+1
+ BRt
t+10n2,t+1
, and Rt = F
X1,tEtX1,t+1
,
with X1,t+1 = [t+1 Rt bt+1 V ATt r
nt b
ft+1
], and EtX2,t+1 = [Etxt+1 Ett+1]. The
matrices A0, A1 and B are given by:
A0 =
266666666666664
1 0 0 0 0 0 0 0
0 1 0 0 0 0 0 0
0 0 1 0 0 G(oY 1)1
2 0 0
0 0 1 1 0 0 0 0
(1 oY )1
(1
B 2) 0 0 0 1 (1 oY)
1
2(1
B 1 1) 0
0 0 0 0 0 (1 + G
2 + G
(oY
1)2) 0 0
0 0 0 0 C((1 GoY ))1 0 1 C((1 GoY))
1
0 0 0 0 0 2 0
377777777777775
,
where B 1 + G2 + G(oY 1)(1/)2.
A1 =
2666666666666664
0 0 0 0 0 0 0
0 0 0 0 0 0 0 0
G1 (oY 1) 0
1G 0 0 G(oY 1) 0
2 0 0 0 0 0 0 0
(1 oY )1 0 0 0 0 0 0 0
G(oY 1)1
0 0 0 0 1 0 0
0 0 0 0 0 0 1 0
0 0 0 0 0 1
3777777777777775
,
B =
0
1
0
0
0
0
c((1 GoY))1
,
F = [0 0 0 0 0 (1 ) (1 )x].
23
8/4/2019 Mayer, Grimm - Counter Cyclical Fiscal Policy, Simple Rules and Price Dispersion
24/26
References
[1] Altig, D., L. J. Christiano, M. Eichenbaum, and J. Linde (2005): Firm-Specific
Capital, Nominal Rigidities and the Business Cycle, NBER Working Paper No. 11034.
[2] Alvarez, L. J., E. Dhyne, M. Hoeberichts, C. Kwapil, H. Le Bihan, P. Launnemann,
F. Martins, R. Sabbatini, H. Stahl, P. Vermeulen and J. Vilmunen (2006): Sticky
Prices in the Euro Area: A Summary of New Micro-Evidence, Journal of the European
Economic Association, 4(2-3), 575-584.
[3] Calvo, Guillermo A. (1983): Staggered Prices in a Utility-maximizing Framework, Jour-
nal of Monetary Economics, Vol. 12, 383-398.
[4] Canzoneri, M., R. Cumbi and B. Diba (2007): The Cost of Nominal Rigidity in NNS
Models, Journal of Money, Credit and Banking, 39(7), 1563-1586.
[5] Canzoneri, M., R. Cumbi and B. Diba (2003): New Views on the Transatlantic Trans-
mission of Fiscal Policy and Macroeconomic Policy Coordination, in Marco Buti (ed.)
Monetary and Fiscal Policies in EMU: Interactions and Coordination, Cambridge Uni-
versity Press.
[6] Coenen, G; Mc Adam, P. and R. Straub (2008): Tax Reform and Labour-Market
Performance in the Euro Area: A Simulation-Based Analysis Using the New Area-Wide
Model, Journal of Economic Dynamics and Control, forthcoming.
[7] Clarida R., J. Gali and M. Gertler (1999): The Science of Monetary Policy: A New
Keynesian Perspective, Journal of Economic Literature, 37, 1661-1707.
[8] Del Negro, M., F. Schorfheide, F. Smets, and R. Wouters (2005): On the Fit and
Forecasting Performance of New Keynesian Models, ECB Working Paper Series, No. 491.
[9] Erceg, C., D. Henderson and A. Levin (2000): Optimal Monetary Policy with Stag-
gered Wage and Price Contracts, Journal of Monetary Economics, 46, 281-313.
[10] Gali J., M. Gertler and D. Lopez-Salido (2007): Markups, Gaps, and the Welfare
Costs of Business Fluctuations, The Review of Economics and Statistics, 89(1), 44-59.
[11] Gali J., M. Gertler and D. Lopez-Salido (2001): European Inflation Dynamics,
European Economic Review, 45, 1237-1270.
[12] Gali J., D. Lopez-Salido and J. Valles (2007): Understanding the Effects of Govern-
ment Spending on Consumption, Journal of the European Economic Association, 5(1).
227-270.
24
8/4/2019 Mayer, Grimm - Counter Cyclical Fiscal Policy, Simple Rules and Price Dispersion
25/26
[13] Gali J. and, D. Lopez-Salido, and J. Valles (2004): Rule of thumb Consumers and
the Design of Interest Rate Rules, Journal of Money Credit and Banking, 36(4), 739-764.
[14] Gali J. and T. Monacelli (2008): Optimal Monetary and Fiscal Policy in a Currency
Union, Working paper, First draft October 2005.
[15] Gali J. and T. Monacelli (2005): Monetary Policy and Exchange Rate Volatility in a
Small Open Economy, Review of Economic Studies, 72, 707-734.
[16] Linnemann, L. and A. Schabert (2003): Fiscal p olicy in the new neoclassical synthesis
Journal of Money, Credit, and Banking, 35, 911-929.
[17] Leith, C., and J. Malley (2005): Estimated General Equilibrium Models for the Evalua-
tion of Monetary Policy in the US and Europe, European Economic Review, 49, 2137-2159.
[18] Leith, C. and S. Wren-Lewis (2007): Counter Cyclical Fiscal Policy: Which Instrument
is Best?, Working paper.
[19] Rabanal, P., and J. F. Rubio-Ramirez (2003): Comparing New Keynesian Models in
the Euro Area: A Bayesian Approach, Working Paper 2003-30, Federal Reserve Bank of
Atlanta.
[20] Schmidt-Grohe S. and M. Uribe (2006): Optimal Fiscal and Monetary policy in a
medium scale macroeconomic model, ECB Working Paper Series, No. 612.
[21] Schmidt-Grohe S. and M. Uribe (1997): Balanced-Budget Rules, Distortionary Taxes,
and Aggregate Instability, Journal of Political Economy, 105, 976-1000.
[22] Smets, F. and R. Wouters (2003): An Estimated Stochastic Dynamic General Equi-
librium Model of the Euro Area, Journal of the European Economic Association, 1(5),
1123-1175.
[23] Soderlind, P. (1999): Solution and Estimation of RE Macromodels with Optimal Policy,
European Economic Review, 43, 813-823.
[24] Taylor, J. B. (2000): Reassessing Discretionary Fiscal Policy, Journal of Economic
Perspectives, 14, 21-36.
[25] Tinbergen, J. (1959): Economic Policy: Principles and Design, The Economic Journal,
69(274), 353-356.
[26] Welz, P. (2006): Assessing Predetermined Expectations in the Standard Sticky Price
Model: A Bayesian Approach, ECB Working Paper Series, No. 621.
25
8/4/2019 Mayer, Grimm - Counter Cyclical Fiscal Policy, Simple Rules and Price Dispersion
26/26
[27] Woodford, M. (2003): Interest and Prices. Foundations of a Theory of Monetary Policy,
Princeton University Press, Princeton.
26
Top Related